What is the recommended stepwise imaging workup for abnormal uterine bleeding?

Advanced Imaging Workup for Abnormal Uterine Bleeding

Start with combined transvaginal and transabdominal ultrasound as your initial imaging—this is the only appropriate first-line test, and CT or MRI have no role in initial evaluation 1.

Initial Imaging: Transvaginal Ultrasound (TVUS)

Transvaginal ultrasound combined with transabdominal approach is the mandatory first step 1. The transabdominal component is essential when the uterus is significantly enlarged or when TVUS has limited field of view 1.

Key Technical Points:

• Doppler imaging should be standard, not optional—it helps identify vascular pedicles in polyps (specificity 62-98%) and improves adenomyosis diagnosis 1
• For postmenopausal women: endometrial thickness ≤4 mm has nearly 100% negative predictive value for cancer 1
• For premenopausal women: there is no validated endometrial thickness cutoff—thickness is NOT an indicator of pathology, and even <5 mm can harbor polyps or other lesions 1
• Thickened endometrium ≥5 mm in postmenopausal women mandates endometrial tissue sampling 1

Diagnostic Performance:

TVUS detects adenomyosis with 82.5% sensitivity and 84.6% specificity, though this drops significantly (sensitivity 33.3%) when coexisting fibroids are present 1. Recent evidence shows that TVS alone provides optimal endometrial assessment in only 63% of cases, with adenomyosis and unfavorable uterine position being the most common limiting factors 2.

Second-Line Imaging: Sonohysterography

If TVUS shows focal endometrial abnormality or is suboptimal, proceed immediately to saline infusion sonohysterography 1. This is not optional when focal lesions are suspected.

When to Use:

• Focal endometrial abnormality on initial TVUS
• Thickened endometrium in postmenopausal women to distinguish focal vs. diffuse pathology 1
• Suboptimal TVUS due to adenomyosis, unfavorable uterine position, or fibroids 2

Diagnostic Capabilities:

• 97% accuracy distinguishing polyps from leiomyomas 1
• Identifies specific features: intact myometrial-endometrial interface + single vessel + acute angle = polyp; absent interface + arborized vessels + obtuse angle = submucosal fibroid 1

Critical Limitation:

Sonohysterography CANNOT reliably distinguish benign from malignant endometrial pathology—endometrial sampling or hysteroscopy remains mandatory when pathology is identified 1.

Third-Line Imaging: MRI Pelvis

Reserve MRI for when ultrasound is inconclusive, indeterminate, or the uterus cannot be fully visualized 1. MRI has no role as initial imaging 1.

Specific Indications:

• Uterus incompletely visualized by ultrasound
• Indeterminate findings requiring better tissue characterization
• Pre-treatment assessment of leiomyomas (location, size, excluding leiomyosarcoma) 1
• Suspected malignancy requiring staging information

Technical Requirements:

• Gadolinium-based IV contrast is preferred 1
• Diffusion-weighted sequences should be strongly considered 1

Diagnostic Performance:

• Endometrial cancer: 79% sensitivity, 89% specificity 1
• Leiomyosarcoma: 100% sensitivity and specificity 1
• Adenomyosis: 78% sensitivity, 93% specificity 1
• Abnormal diffusion-weighted signal + irregular endometrial-myometrial interface best differentiates benign from malignant (AUC 0.89) 1

Follow-Up Imaging Scenarios

If Initial Workup is Negative:

Repeat TVUS to reassess the endometrium—cancers can be missed on initial imaging or sampling 1. If endometrium remains ≤4 mm in postmenopausal women, negative predictive value for cancer is nearly 100% 1.

If Surveillance is Appropriate:

• MRI with gadolinium and diffusion-weighted imaging for monitoring leiomyomas or suspected malignancy 1
• Repeat TVUS for benign conditions
• CT has no role in follow-up imaging 1

Common Pitfalls to Avoid

1. Never use CT pelvis for AUB evaluation—there is no supporting literature for its use in initial, follow-up, or surveillance imaging 1

2. Don't rely on endometrial thickness alone in premenopausal women—assess echogenicity and texture instead 1

3. Don't skip the multimodal approach when TVS is suboptimal—using transabdominal, transrectal, or saline contrast can optimize the angle of insonation and avoid unnecessary invasive procedures in 30.5% of cases 2

4. Don't assume sonohysterography can exclude cancer—it cannot, and tissue diagnosis remains essential 1

5. Don't order MRI first—it's expensive, time-consuming, and ultrasound provides equivalent or better information for initial triage 1

Integration with Tissue Diagnosis

When endometrial tissue assessment is indicated, sampling should be performed in conjunction with hysteroscopic evaluation, with visually directed biopsy favored over blind sampling 3. Hysteroscopy demonstrates higher diagnostic accuracy than blind endometrial biopsy for detecting endometrial cancer, hyperplasia, polyps, and submucosal myomas 3.